actuallyalys

It feels like it’s probably most applicable to people who do similar work as you and who you talk shop with. I wouldn’t consider these work friends, which to me are people at the same place of work that you wouldn’t hang out with if not for work.

I daresay ClojureScript typically has a significant influence, but it's more social than technical. Many of the examples and tutorials are going to favor a certain style, and people who choose ClojureScript mostly do so because they like Clojure.

But if you want to do tons of JavaScript interop, you can. If you want to adopt some syntax, you can (through macros). If you want to use jQuery directly, you can. If you want to use a React wrapper, you can. There's not really technical limits pushing you toward a style or architecture, although you still have the underlying limits of JavaScript and the runtime environment.

I probably shouldn’t have said “standard,” but in my experience, they do (usually!) follow certain conventions. They’re not formal conventions, sure, but they are conventions.

In my experience, short identifiers in Clojure tend to be pretty standard things, e.g., f for a function, coll for a collection. Certainly this article has some pretty cryptic examples, but the author agrees with you and me and ends up fixing them.

Carmack is right to some extent, although I think it’s also worth mentioning that people replace their computers for reasons other than performance, especially smartphones. Improvements in other components, damage, marketing, and status are other reasons.

It’s not that uncommon for people to replace their phone after two years, and as someone who’s typically bought phones that are good but not top-of-the-line, I’m skeptical all of those people’s phones are getting bogged down by slow software.

I would like to know why Boeing didn’t have a replacement ready when the 757 stopped being produced. I realize commercial planes are not consumer goods, but 2004 still seems like a long time ago. Thus I imagine having a 757 replacement sooner was the plan, so what derailed it?

Limits on written records and the limits of what we can derive from genetic analysis means 536 and other years these analyses uncover are probably best understood as local minima rather than definitively the worst.

I suppose it sorta makes sense, if you think of it as a free mobile game model, where the goal is to make more money by get more people to pay (and especially by getting the whales to pay more) and to do that you need new daily or weekly content. Although, I feel like that model scales just fine with humans creating the content.

> “It helps us get closer to our mission. To teach well, we need to create a massive amount of content, and doing that manually doesn’t scale. One of the best decisions we made recently was replacing a slow, manual content creation process with one powered by AI. Without AI, it would take us decades to scale our content to more learners. We owe it to our learners to get them this content ASAP.”

It’s hard to fathom why this would be the case. Isn’t creating learning materials mostly an upfront task? Obviously you want to update your materials and fix mistakes. And perhaps you sometimes have a new idea that necessitates an update or rewriting things. But creating lesson materials honestly seems like it would scale very well.

I think about reading language learning blogs and forums and people would sometimes recommend favorite resources that were from decades ago. I’m sure something was lost in them not being totally up to date but honestly, persistence is way more important than currency, let alone “having a massive amount of content.”

I read the original post as saying that no one has pushed the engine to the extent a completed AAA game would in order to uncover performance issues, not that performance is bad or that Bevy devs haven’t worked hard on it.

Fasterthanlime recommended it a while back on Mastodon so that did pique my interest. The focus on AI kind of put me off Zed, though. I don’t know whether that’s fair—I see one of their blog posts touts “Out of Your Face AI”—but that was my reaction.

I should honestly give DAP a try in Neovim. I think DAP being newer, seemingly getting less attention than other elaborate technologies to integrate like LSP or Treesitter, and having a more complex UI made me suspect it might not be reliable. Adding a DAP plugin to a secondary editor also feels less risky. But it sounds like my concerns may have been addressed, if not overblown all along.

I don’t think it will dethrone Neovim for me, but this makes me wonder whether Kate could become my second editor and allow me to largely drop VS Code, especially with the DAP support. The session support also looks interesting.

It wouldn’t shock me if multimodal LLMs were good at GeoGuesser [0], but if we’re being picky, it takes more than a few examples to demonstrate a game is “solved.” I also wonder what kind of data leakage might have been at play, like other people have suggested.

To be clear, my point is not that this is unimpressive, just that this doesn’t demonstrate much. (Edit: I should have said, it doesn’t demonstrate what the title claims.)

[0] they were very likely trained on a large number of photos that had their location, and they have the ability to isolate features. Combinined with their ability to interpret instructions and just, well, guess, that seems like you have enough for the game.

Depending on how work was divided up, it doesn’t seem like the photographer of Jack Nicholson would necessarily know where the image Nicholson was superimposed on came from, so I don’t blame them for not checking with him first.

Sure, but AWS has human customer support and engineers overseeing all that (not to mention accountants and lawyers).

The only way I could see this not ending badly is if you’re renting space for your friends. Presumably you would be selling whatever you’re renting at cost.

Even renting space within a community you’re a part of, like a local makerspace, seems like a bad idea. While it would probably mitigate the worst risks, could still leave you with people dissatisfied with the level of service.

Thanks for the fact check. Saying you can get even better than 4K is definitely overstating it.

It’s possible the sources I was looking at were thinking of later eras of film, which I’d assume would have higher quality (and the lenses certainly improved).

Thanks for sharing that! I’m sure there’s a lot of nuance. I was kinda shocked at the long list of motion picture film formats listed on Wikipedia when I was trying to figure out the equivalent resolution 35mm film has.

I’m sure sock puppets and voting rings play a role, but “this link was interesting to a bunch of people on sites with some overlap in interests and demographics” seems a simpler and better explanation.

This article is kind of strange to me because it implies celluloid film is low quality, but it seems there’s a rough consensus that you can can scan that size of film in to create good 4K video, if not higher. That’s not enough for the Sphere’s full 16K resolution, but that’s more to do with it having extremely high resolution than the film being “tiny” or grainy.

This paper [0] suggests improvements in car emissions has played a big role in reducing emissions in European cities as a whole. Vehicle emissions of all kinds have fallen pretty dramatically across Europe [1], although this is total emissions for vehicles, so it includes policies to reduce driving as well as those to reduce each vehicle’s emissions. So overall trends toward more efficient cars are certainly part of the story. Given these images are between 2007, when emissions had already been falling, and 2024, I’m inclined to think the policies highlighted in the article played a significant role as well.

[0]: https://www.sciencedirect.com/science/article/pii/S259016211...

[1]: https://www.eea.europa.eu/en/analysis/indicators/emissions-o...

I mean, maybe he deserved to get into more universities, but he did get into solid universities, even if they’re not at the very top. The application pool is very competitive at top universities, and I’m not sure business success should be an automatic in.

It’s a pretty strange choice by Microsoft, which historically hasn’t been afraid to have multiple products that are similar to one another, often to the point of confusion.

DOGE is not merely “recommending areas where fraud/waste/abuse” can be cut:

Firing and then trying to rehire nuclear safety employees: https://www.bbc.com/news/articles/c4g3nrx1dq5o.amp

Firing employees at the CDC and other health agencies: https://www.npr.org/sections/shots-health-news/2025/02/17/nx...

(But not the “disease detectives” as originally reported: https://www.cbsnews.com/amp/news/cdc-disease-detectives-doge...)

Laying off IRS employees: https://www.axios.com/2025/02/20/irs-layoffs-trump-doge

A list of cuts and firings: https://federalnewsnetwork.com/agency-oversight/2025/02/a-co...

Additionally, many of the claims by Musk and others being made about the agencies being cut are false: https://www.forbes.com/sites/conormurray/2025/02/19/here-are...

It is a bit funny the era he compares today to is 1892—the gilded age.

Anyway, 1982 may not be a very appealing low-Gini coefficient period to return to, but 1892 also seems like a bad time. I actually found the comparison interesting enough, even though the arguments fall flat, but I think we need to look more broadly to understand inequality.

This is what my brother and I did too. (My sister was young enough to mostly experience the "entropy box.") If I had to guess, I'd imagine a plurality of lego players do it that way, if not a majority, with smaller groups of people exclusively building from instructions or rarely building from instructions.

It does mention doing load shedding:

> By default searchcode prioritizes system survival (hey its a free service!), and as such might do some load shedding, which can mean you don't see results you expect. You can do some things to help with this.

I don’t know whether that explains the results. Some similar queries, e.g., searching for `math.ceiling`, do return many results.

Funnily enough, my impression is that Clojure, the most popular Lisp on the JVM, creates a lot of garbage—probably more than Java—but I believe that’s due to it using immutable collections and not due to it being a Lisp.

Unfortunately, I can't access the full text of the study directly, but I would argue it would be more accurate to say it impacted (past tense) 1 million children, assuming this paragraph from the news story is correct:

> Results of the analysis, published in the journal JAMA Pediatrics, showed approximately 1.01 million children, or 1.4%, are believed to have ever experienced long COVID as of 2023 and about 293,000, or 0.4%, were experiencing the condition when the survey was being conducted.

There are certainly reasons to think it might be under-reported, so perhaps that should go in the headline (or subheadline) as well. But I don't think the headline should suggest this particular study estimated 1 million children were experiencing it at the time.